Food environments are shaped by the spatial distribution of opportunities to purchase healthy and affordable food, and limited access to these opportunities contributes to health inequities. Evidence remains limited, however, on how online grocery services are reshaping food accessibility. This study examines Seoul's food environment by measuring spatial accessibility to both offline and online grocery stores using the enhanced 2-step floating catchment area (E2SFCA) method and an integrated accessibility index using a weight-based modified E2SFCA approach. Results show that the advent of online grocery stores does not alleviate low offline food accessibility; instead, low accessibility persists in Seoul's peripheral neighborhoods, suggesting that online grocery services might reinforce existing spatial disparities where traditional food deserts are concentrated. By considering both offline and online grocery stores, this study extends food accessibility research to account for online grocery shopping opportunities and provides empirical evidence to inform nutrition policies.
The article examines the relationship between the energy sector, safety, and the militarization of social and environmental space. It advances the thesis that centralized energy systems can constitute part of, and provide critical infrastructure for, the modern military apparatus; they are both potential targets of attack and integral components of state and military security. The scale and location of energy sources, as well as their forms of management and ownership, influence the weaponization of energy systems. In this context, decentralized energy production, distribution, and governance, alongside a shift to local energy sources as part of the energy transition, represent the most effective guarantees for the peaceful use of energy. Decentralization of the energy system based on renewable sources not only facilitates the implementation of climate policy but also strengthens energy security and contributes to the demilitarization of energy. The experience of the April 2025 lockdown in Spain, combined with the war in Ukraine, illustrates that large, centralized energy infrastructures are less resilient and less secure in times of disruption or crisis than decentralized systems. From this perspective, large coal-fired and nuclear power plants, as well as centralized forms of energy distribution, pose risks to both energy and military security. Consequently, the choice of technologies underpinning the energy system is of fundamental importance for environmental security, societal safety, and military security.
This article offers a methodological reflection on adopting a particular type of ethnography-event ethnography-to engage with the production and circulation of policy realities. Responding to recent geographical calls asking us to critically investigate the Blue Economy (BE)-a recent policy paradigm supposedly ensuring economic development, social well-being, and environmental sustainability-I detail my choice of event ethnography as the primary methodological choice informing my encounter with one such paradigm in the archipelagic state of the Republic of Seychelles. In doing so, this article expands on current qualitative methodological approaches for the study of the BE by foregrounding the potentialities of event ethnography as a generative approach to engage with "policies in-the-making." Ultimately, I posit that event ethnography provides an additional and unique methodological entry point to presently dominant normative and prescriptive analyses of BE projects. Simultaneously, and beyond the BE as a research problematic per se, I suggest that event ethnography affords geographers with rare spatiotemporal configurations to productively engage with the messiness and heterogeneity of policy realities-be those global in reach, or unfolding in more localized scales-especially when those realities are in-the-making.
Behavioral change toward low-carbon lifestyles is a key strategy for addressing climate change. Individuals encounter numerous barriers to such change, however, one of which is the psychological distance of climate change-where it is perceived as abstract, distant, and irrelevant. As attempts to reduce psychological distance have produced mixed results, bridging this distance-enabling individuals to perceive and relate to climate change as a concrete and connected issue-could offer a more effective approach. This study examines the role of geography education in addressing these challenges. Using survey data from the Taiwanese public (N = 579), the study investigates whether a positive experience of learning geography promotes climate change mitigation directly, and whether it does so indirectly through cultivating the sense of global connectedness, which helps bridge the psychological distance of climate change. The results indicate that the experience of learning geography is positively associated with the sense of global connectedness. Moreover, global connectedness partially mediates the relationship between the experience of learning geography and behavioral motivation. These findings suggest that providing quality geography education can be an effective pathway to addressing complex climate change challenges, including behavioral change and psychological distance.
Travel intentions triggered by mortality salience are shaped by a complex interplay of psychological and environmental factors that might either encourage or constrain travel behavior. This interdisciplinary study adopts a geo-psychological framework that integrates psychological factors (e.g., self-esteem and cultural worldview) with environmental factors (e.g., COVID-19 severity and tourism cluster) to investigate the spatial variability of travel intentions among U.S. residents facing heightened mortality awareness during the COVID-19 pandemic. Using a combination of primary and secondary data and applying spatial analytical methods, this study demonstrates that psychological and environmental factors are associated with mortality salience-driven travel intentions, and that these effects vary significantly across geographic contexts. By incorporating spatial heterogeneity into psychological processes, this geo-psychological perspective offers a novel framework for understanding how tourists perceive, decide, and behave within and between destinations.
This study examines the spatial dimensions of college football recruiting by quantifying Euclidean distance (pull power) and concave hull-based recruiting areas (recruiting footprints) across all sixty-four Power 5 football programs from 2011 to 2020. Using geographic information systems, the hometowns of 13,554 recruits were geocoded to calculate aggregate recruiting distances and spatial footprints for each program. Informed by concepts of spatial interaction, retail gravitation, and athletic departments' pull power, this study introduces the recruiting footprint as a concave hull-based spatial measure that captures both the breadth and configuration of a program's recruiting area. Results reveal substantial variation across conferences and institutions, with Pac-12 programs exhibiting both the greatest pull power and the most expansive recruiting footprints. Correlation analyses indicate that institutional resources are positively associated with pull power, whereas broader recruiting footprints are negatively associated with subsequent on-field performance. Together, these findings highlight the distinct roles of distance and geographic dispersion in recruiting strategy and establish a spatial baseline for examining athlete migration and recruiting behavior in an increasingly mobile and geographically reconfigured college football landscape.
Technological innovation patterns are fundamental to understanding regional and industrial innovation trajectories. This study proposes a two-dimensional conceptual framework integrating technology sources and attributes to classify regional innovation patterns. Taking the Yangtze River Economic Belt as a case, we examine the characteristics and territorial correlations of these patterns. Results show that the endogenous-generic pattern is concentrated in developed cities, driven by government investment and human capital toward frontier breakthroughs. The endogenous-specialized pattern emerges in subcentral cities, characterized by independent innovation in specific domains. The exogenous-generic pattern suits developing regions pursuing catch-up through technology introduction and local adaptation. The exogenous-specialized pattern prevails in less developed areas, where specialized applications are built on existing industrial bases via external resource integration. These differentiated pathways reflect variations in factor endowments and socioeconomic conditions, with key determinants including government involvement, financial support, human capital, transportation infrastructure, and industrial structure shaping pattern selection. This study extends existing innovation pattern classifications, deepens geographic understanding of regional innovation disparities, and offers an integrated analytical framework for place-based strategies across different territorial contexts.
In this article, we examine the experiences of five undergraduate students who enrolled in an introductory drones class to learn how to operate drones, gather geospatial information, and analyze the data generated through student flights. We employed open-ended surveys to gather data from the students about their expectations, experiences, and reflections from the class. Our analysis focused on three primary objectives: (1) an examination of students' prior experience with drones and how that affected their ability to learn new concepts surrounding drones; (2) an examination of students' expectations before coming into the class to work with drones; and (3) an examination of students' ease of working with drone data and their understanding of what can be achieved with the data that were derived from drones. Our results indicate that student experiences were improved through repetition to develop technical and technological fluency, a challenge-based approach to learning created opportunities for students to overcome unexpected challenges, and small class sizes that allow for more frequent and meaningful connections between students and instructors. Based on these findings, we argue instructors should keep classes small, build a schedule that fosters repeated engagement, and embrace a challenge-based learning approach to the class.
This article applies geographical sequence analysis (GSA) as a powerful toolkit for researching evolutionary dynamics in economic geography. Although entrepreneurial ecosystems (EEs) are known to be dynamic, conventional methods often fail to capture their internal, microgeographical evolution over time. Applying GSA to a longitudinal data set of startup locations in Berlin, the evolutionary trajectories of raster cells within the Berlin EE are classified. The analysis uncovers a clear "evolutionary mosaic" of three distinct trajectory types-established cores, secondary centers, and young growth zones. This study contributes by both offering a new, place-sensitive conceptualization of EEs as evolving mosaics and by demonstrating a methodological framework for analyzing complex spatiotemporal change.
Interactive dashboards have become central tools for visualizing policy-relevant data, but they remain limited by rigid structures and technical complexity. Recent advances in large language model (LLM)-powered conversational AI offer a way to address these challenges, as LLMs can interpret unstructured natural language inputs and translate them into structured functions. This proof-of-concept study introduces a Generative AI (GenAI)-powered chatbot that integrates LLMs with an online dashboard, demonstrated through the India Policy Insights (IPI) dashboard. The system translates conversational queries into structured functions, retrieves validated outputs from a spatially enabled database, and presents results as text, charts, and maps. We implemented 13 representative functions spanning spatial, temporal, composite, classification, and constraint-based analyses. Results show that spatial, classification, and constraint-based functions achieved consistently high accuracy due to explicit parameters, while composite multi-indicator functions posed greater challenges. The findings demonstrate the potential of GenAI-powered interactive dashboards to support language-driven interaction and broaden accessibility for diverse user groups, regardless of technological literacy. Beyond the case study, the proposed framework provides a design pathway for developing GenAI-powered dashboards that democratize access to spatiotemporal data, enhance evidence-based policymaking, and help bridge persistent gaps between data and actionable insights.
There is a growing need to prepare students to address complex environmental challenges, particularly climate change and the expanding role of artificial intelligence (AI). This article examines the adaptation of the Artificial Intelligence for Social Good (AI4SG) Ideation module, a pitch- and project-based teaching framework designed to integrate AI literacy and sustainability education. Implemented in an upper division climate change course, the module combines project-based learning, culturally responsive pedagogy, and geographic thinking to guide students in developing community-oriented AI proposals aligned with the United Nations Sustainable Development Goals (SDGs). Students were encouraged to approach their communities not through deficit lenses, but through additive lenses, and to frame sustainability challenges across campus, local, and metropolitan scales. To examine student engagement with the module, we analyzed lab reports, team proposals and pitches, and end-of-semester reflections. Findings indicate that teams articulated SDGs frameworks at local scales and conceptualized AI applications grounded in community context, such as consumption-related carbon emissions and climate-related disaster response. Student reflections further suggest developing critical AI literacy, including attention to ethical considerations and the limitations of technological solutions. By situating AI within applied geographic reasoning and pitch-based pedagogy, the AI4SG Ideation module provides an example of how environmental and technological literacy can be integrated within undergraduate geography education.
The history of geography is a history of geographical practices. In our book Weltbildwechsel (Schlottmann and Wintzer 2019), literally "shift of worldviews," we propose to tell the history of the discipline by considering the manifold ways of engaging with the world: measuring, explaining, conquering, teaching, enlightening, perceiving, designing, differentiating, visualizing, and modeling. Each of these practices denotes a specific way of doing geography, embedded in its historical, social, and epistemic contexts. Rather than offering one coherent, linear, and continuous narrative, this approach highlights the existence of different worldviews enacted within and over time, uncovering simultaneity in the nonsimultaneity. In this commentary we revisit our approach by considering the recent contribution by Kinkaid et al. (2025). We argue that such a historiography opens new pedagogical and epistemological avenues for teaching geography's histories as plural, entangled, contingent, and at any time practice-based.
Charitable food assistance is vitally important for many food-insecure households, but relatively little research has been done to assess the availability of these programs on a broad scale. This article describes a project to analyze the spatial availability of these programs in the state of Georgia. We collected data on charitable food agencies from nine regional food banks in February 2023. We created proximity scores based on three measures of agency proximity and density at the census tract level. Controlling for population density, we identified spatial hot and cold spots for agency proximity and identified distinct demographic characteristics for these areas. Our analysis identifies hot spots for charitable food availability in areas of high poverty concentration in the metropolitan Atlanta area and sections of south Georgia. Cold spots are found in the northern Atlanta area and other small sections of the state. This analysis demonstrates how food-insecure households in mixed-income urban areas might have difficulty accessing charitable food programs. Our findings correspond with other research showing how suburban and other mixed-income areas can amplify economic precarity. This project provides a model of how broad-scale research on charitable food can be conducted through collaborative research.
Crowdsensing has become an increasingly important source of socioenvironmental data. The locations collected through these networks also raise geoprivacy concerns, however. This study provides one of the first large-scale empirical examinations of location masking, a common privacy protection behavior where the reported sensor location is intentionally displaced from the true location, with the aim to address a current critical gap in understanding how such behaviors occur "in the wild" within large-scale, real-world crowdsensing networks. We leverage a large national data set of PurpleAir sensors and apply ordinal logistic and mixed-effects models to examine how the immediate sensor placement and broader neighborhood characteristics shape location masking. Our results show that sensors placed indoors, in nonurban, or in nonresidential areas tend to exhibit high levels of masking, meaning their reported locations deviate further from the true locations. In contrast, sensors in neighborhoods with high educational attainment, income levels, older populations, and larger proportions of non-white and Hispanic residents are associated with low levels of location masking. These findings indicate the importance of considering both the physical sensor placement and sociospatial context in shaping privacy-related behaviors and suggest that such factors should be carefully considered when designing and promoting crowdsensing initiatives.
Understanding where and how people perceive infection risk is important for effective public health communication. This study examined COVID-19 risk perceptions among young adults in Seoul and Tokyo. Using a qualitative GIS approach, we identified the geographic distribution of perceived risk and the reasons behind these perceptions. The analysis revealed that high-risk perceptions were triggered by overcrowding and were concentrated in city centers. By contrast, low-risk perceptions tended to be dispersed and were shaped by lower population density and individuals' place attachment. One counterintuitive finding was that open spaces, which are generally considered safe due to their spaciousness, may paradoxically relax mask-wearing norms and thereby heighten fear of infection among others. These findings suggest that perceived infection risk is co-constituted by environmental factors, individual experiences, and social norms in urban spaces.
Natural language processing (NLP) applications for extracting spatial information from text data require robust, reproducible, and trustworthy modeling approaches. Our study enlisted a latent Dirichlet allocation algorithm to extract latent spatial patterns in published scientific literature about the Mississippi River Basin. We focused on methods to address subjectivity and refine three "geospatial topic modeling" steps: data preparation, model calibration, and output interpretation. First, a nomenclature that categorizes text data by spatial extent mitigates risk of unexpected geographies and concepts obscuring latent topic interpretation. Second, sensitivity analysis allows for individual effects of geospatial topic model parameters to be identified, thus informing the calibration and removal of parameters necessary to refine latent topic meaningfulness. Finally, "junk topics" constructed by topic models convey language that distinguishes latent topics that possess geographical importance, thus contributing to interpretation of meaningful spatial patterns. These approaches allow us to address and mitigate subjectivity in the application of NLP for identifying spatial patterns in text data, through a combination of high-level statistics, sensitivity analysis, and human interpretation. Such techniques are applicable beyond geospatial analysis of the Mississippi River Basin's environmental challenges, particularly as the demand for high-quality text data in NLP-based geographical studies grows in the coming years.
Central America and Nicaragua are frequently affected by disasters due to their physical and environmental characteristics, exposure, and vulnerability. This study begins by offering a physical-geographic characterization of Nicaragua, followed by a national-scale analysis using the EM-DAT database to examine the most intense disasters from 1930 to 2023. We then focus on a municipal-level analysis using the available DesInventar database between 1992 and 2013 to identify municipalities with the highest disaster incidence. We filled the last decade (2014-2023) with available information extracted from secondary sources such as large events reports and scientific papers. Key findings highlight the most impactful years, months, and types of disasters, such as floods, tropical cyclones, droughts, wildfires, and epidemics. Subsequently, statistical analyses are conducted, including a correlation matrix, multiple linear regression, Poisson regression, and random forest modeling. These methods help determine which socioeconomic variables best explain the occurrence of disasters by municipality throughout history. Our results are crucial for understanding future disaster risk, explaining the spatiotemporal patterns, and identifying disaster hot spots. This study provides valuable insights into disaster risk management and serves as a methodological example for countries with similar conditions-tropical regions and developing nations that often lack comprehensive baseline data.
